Marloes Nies*, Mariëlle S. Klein Hesselink*, Gea A. Huizinga, Esther Sulkers, Adrienne H. Brouwers, Johannes G.M. Burgerhof, Eveline W.C.M. van Dam, Bas Havekes, Marry M. van den Heuvel-Eibrink, Eleonora P.M. Corssmit, Leontien C.M. Kremer, Romana T. Netea-Maier, Heleen J.H. van der Pal, Robin P. Peeters, John T.M. Plukker, Cécile M. Ronckers, Hanneke M. van Santen, Wim J.E. Tissing, Thera P. Links, and Gianni Bocca
*These authors contributed equally to this study.
J Clin Endocrinol Metab. 2017, 102: 1218–1226.
7
ABSTRACT
Context
Little is known about long-term quality of life (QoL) of survivors of pediatric differentiated thyroid carcinoma (DTC). Therefore, this study aimed to evaluate generic health-related QoL (HRQoL), fatigue, anxiety, and depression in these survivors compared with matched controls, and to evaluate thyroid cancer-specific HRQoL in survivors only.
Design
Survivors diagnosed between 1970 and 2013 at age ≤18 years, were included. Exclusion criteria were a follow-up <5 years, attained age <18 years, or diagnosis of DTC as a second malignant neoplasm. Controls were matched by age, sex, and socioeconomic status. Survivors and controls were asked to complete 3 questionnaires [Short-Form 36 (HRQoL), Multidimensional Fatigue Inventory 20 (fatigue), and Hospital Anxiety and Depression scale (anxiety/depression)]. Survivors completed a thyroid cancer–
specific HRQoL questionnaire.
Results
Sixty-seven survivors and 56 controls. Median age of survivors at evaluation was 34.2 years (range 18.8 to 61.7 years). Median follow-up was 17.8 years (range 5.0 to 44.7 years). On most QoL subscales, scores of survivors and controls did not differ significantly. However, survivors had more physical problems (P = 0.031), role limitations due to physical problems (P = 0.021), and mental fatigue (P = 0.016) than controls. Some thyroid cancer-specific complaints (e.g. sensory complaints and chilliness) were present in survivors. Unemployment and more extensive disease or treatment characteristics were most frequently associated with worse QoL.
Conclusions
Overall, long-term QoL in survivors of pediatric DTC was normal. Survivors experienced mild impairment of QoL in some domains (physical problems, mental fatigue, and various thyroid cancer-specific complaints). Factors possibly affecting QoL need further exploration.
7
INTRODUCTION
The diagnosis and treatment of childhood cancer constitute a major life event that may influence psychosocial functioning and quality of life (QoL) long after initial treatment has been completed. Pediatric differentiated thyroid carcinoma (DTC) is a rare malignancy with an excellent survival rate (1, 2). The incidence of this disease is rising (3, 4). At present there is an increased focus on the long-term consequences of this disease and its treatment. After initial treatment, which most often consists of a total thyroidectomy followed by radioactive iodine (131I) administration, life-long follow-up is initiated (5). Moreover, about a third of the pediatric DTC survivors in The Netherlands have postoperative complications, including permanent hypoparathyroidism, which also require medical support (2). Both hypothyroidism and hypoparathyroidism have been described as negatively affecting the QoL of DTC survivors (6, 7).
The available data on QoL in survivors of pediatric DTC are limited. Most studies evaluate QoL in survivors of adolescent and adult DTC. QoL in survivors of adolescent DTC was found to be normal compared to control groups (8, 9). Survivors of adult DTC reported lower health-related QoL (HRQoL) and higher levels of fatigue, anxiety, and depression compared with healthy controls (5, 10-12). For DTC survivors, thyroid cancer-specific complaints may arise after treatment and during follow-up.
Most long-term survivors of adult DTC with a median follow-up period of 9.6 years experienced disease specific symptoms (13). In survivors of pediatric DTC, thyroid cancer-specific complaints have not yet been evaluated.
QoL studies performed in survivors of other childhood cancers showed conflicting results. QoL varied from no differences to lower QoL in survivors, compared to a control or comparison group; this may be explained by the differences in specific clinical characteristics of the different types of cancers in which this has been studied (14-20).
Several specific survivor characteristics may influence QoL during follow-up.
QoL may improve in survivors of adult DTC after a longer time following diagnosis, especially ≥5 years after diagnosis (10, 21). Employment and a higher educational level were found to be associated with better QoL in survivors of adult DTC (12, 21, 22). Young adult survivors reported other thyroid cancer-specific complaints than did older survivors. For example, young adult survivors reported more scar complaints, but less voice complaints (22). QoL in DTC survivors seems to be independent of thyrotropin (TSH) levels (8, 10).
So far, information is limited on long-term QoL in survivors of pediatric DTC.
Therefore, the first aim of our study was to evaluate QoL by means of the self-reported levels of generic HRQoL, fatigue, anxiety, and depression in adult survivors of pediatric DTC compared with those of matched controls. The second aim was to evaluate thyroid cancer-specific HRQoL in adult survivors of pediatric DTC. Our third aim was to evaluate whether long-term generic HRQoL, fatigue, anxiety, depression,
7
and thyroid cancer-specific HRQoL were associated with survivor, tumor, treatment, or follow-up characteristics.
MATERIALS AND METHODS
Survivors
Adult survivors of pediatric DTC who participated in a nationwide follow-up study were asked to participate in this substudy (2). Inclusion criteria for the nationwide study were: diagnosis of DTC between January 1970 and December 2013 at the age of ≤18 years, and treatment in The Netherlands [as described earlier (2)]. Exclusion criteria for this substudy were: follow-up <5 years after diagnosis, attained age <18 years, diagnosis of DTC as a second malignant neoplasm, lack of command of the Dutch language, and
131I administration within 3 months before evaluation.
Controls
To include controls, survivors were asked to approach 1 or 2 persons of similar sex and age (± 5 years of the survivors’ age) at time of follow-up. The only exclusion criterion for controls was having a medical history of malignancy. Preferably, peers were included as controls, trying to match on socioeconomic status. If this was not possible, siblings were allowed to be matched controls.
Participants
All survivors and controls gave informed consent before participating in the study.
The Institutional Review Board of the University Medical Center Groningen approved the study on behalf of all participating institutions.
Medical data
Data on survivor and tumor characteristics (tumor node metastasis classification, histology, and age at diagnosis), treatment (surgical complications, i.e.
hypoparathyroidism and recurrent laryngeal nerve injury, and cumulative 131I dose), and follow-up (follow-up time; outcome: i.e. remission, recurrence or persistent disease; and level of TSH suppression during follow-up) were retrieved from medical records. Data on marital status, job status, and level of education were retrieved from a questionnaire.
QoL assessment
Generic HRQoL, fatigue, anxiety, and depression questionnaires were assessed for survivors and controls. The thyroid cancer-specific HRQoL questionnaire was assessed for survivors only. The questionnaires delineated below were used to assess QoL.
7
Short Form 36
The Short Form 36 (SF-36) (23) is a 36-item health survey, validated for the Dutch population (24). This questionnaire measures self-reported HRQoL on 8 domains:
physical functioning, role limitations due to physical problems, bodily pain, general health, vitality, social functioning, role limitations due to emotional problems and mental health. Scores of domains are presented on a scale from 0 to 100. A higher score indicates better HR-QoL. The scores of the 8 subscales were converted into the physical and mental component scales by means of the algorithm of Ware and Kosinski (25), based on Dutch norms (24). An example of a question in the SF-36 questionnaire is: “In general, would you say your health is…” Possible answers are
“poor,” “fair,” “good,” “very good,” “excellent.”
Multidimensional Fatigue Inventory 20
The Multidimensional Fatigue Inventory 20 (MFI-20) (26) self-report instrument was designed to measure fatigue. Fatigue is scored on 5 domains: general fatigue, physical fatigue, mental fatigue, reduced motivation and reduced activity. The MFI-20 was proven to be valid for the Dutch population (26). The total score, ranging from 20 to 100 (higher scores indicating more fatigue), is calculated as the sum of the 5 domains, and each domain (range 4 to 20) is the sum of 4 items. Each domain includes 4 items on a 1 to 5-point scale ranging from “no, that is not true” to “yes, that is true.” An example of an item is “thinking requires effort.”
Hospital Anxiety and Depression Scale
The Hospital Anxiety and Depression Scale (HADS) (27) is a 14-item self-assessment scale measuring levels of depression and anxiety. This questionnaire has been validated for Dutch subjects (28). Scores on subscales range from 0 to 21. Scores for the total scale can range from 0 to 42. A higher score indicates a higher level of anxiety and/or depression. Answers can be given on a 4-point Likert scale. For example, for
“I feel tense or ‘wound up’,” possible answers are “not at all,” “from time to time, occasionally,” “a lot of the time,” or “most of the time.”
Thyroid cancer-specific HRQoL
The thyroid cancer-specific HRQoL questionnaire (THYCA-QoL) (29) was designed to assess treatment- and cancer-specific side effects of thyroid cancer. For a time frame dealing with the previous week (except in the case of the sexual interest item, which refers to the previous 4 weeks) items can be scored on a 4-point scale ranging from 1 (“not at all”) to 4 (“very much”). Items are summarized and translated into 7 scales (neuromuscular, voice, concentration, sympathetic, throat/mouth, psychological, and sensory problems) and 6 single items (problems with scar, felt chilly, tingling hands/feet, gained weight, headache, interest in sex). Scales range from 0 to 100.
7
A higher score on this scale indicates more complaints. An example of a question in the THYCA-QoL questionnaire is “Have you had a dry mouth?”
Statistical analysis
Characteristics of survivors included in the QoL study were compared to characteristics of the excluded survivors from the nationwide study and to the eligible, but not included, survivors in the QoL study. Characteristics of survivors and controls were compared to evaluate whether controls were a good comparison group for the included survivors. Scores on HRQoL, fatigue, anxiety, and depression of survivors and matched controls were compared. Scores on thyroid cancer-specific HRQoL were analyzed using descriptive statistics. Associations between survivor, tumor, treatment, and follow-up characteristics and scores on QoL were assessed. Single items of the THYCA-QoL were dichotomized (not at all and a little = no; quite a bit and very much = yes) before testing these associations. Comparisons were made using χ2 tests or a Fisher’s exact test (if >20% of the cells had an expected count of <5) in case of categorical variables. Mann-Whitney U and Kruskal-Wallis tests were performed because normality and homogeneity assumptions of dependent variables had been violated. For analyses of possible correlations between 2 variables with ordinal scales, the Spearman rank correlation coefficient was used. The TSH level for each year of follow-up was expressed as the geometric mean of the observed TSH values during that year, as published previously (30).
Missing or unknown values were excluded (listwise deletion) from analyses or, in case of the 4 questionnaires, imputed using questionnaire-specific guidelines. For SF-36 and the THYCA-QoL, scores were imputed when ≤50% values of the scores for this subscale were missing. For MFI-20 and HADS, scores were imputed when ≥2 values of the subscale were not missing. When these criteria were not met, survivors or controls were excluded from analyses of the relevant subscale. IBM SPSS Statistics for Windows version 23 (IBM, Armonk, NY) was used for statistical analyses. Differences were considered statistically significant at P <0.05. To adjust for the multiple testing, a significance level of 1% (P <0.01) was used for the tested associations between survivor, tumor, treatment, and follow-up characteristics and QoL.
RESULTS
Characteristics of survivors of pediatric DTC and matched controls Out of 105 survivors participating in the nationwide study (2), 75 survivors were eligible for this long-term QoL substudy, 67 of whom (89.3%) agreed to participate.
Eight survivors eligible for the QoL study were not included (Figure 1). As expected, survivors of the nationwide study, excluded for the QoL study, and participating survivors in the QoL study differed in age at evaluation (median 19.1 vs. 34.2 years, respectively; P <0.001) and follow-up duration (median 2.8 vs. 17.8 years, respectively.
Figure 1. Flowchart of inclusion of adult survivors of pediatric differentiated thyroid carcinoma in
7
the quality of life study.
Table 1. Characteristics of adult survivors of pediatric differentiated thyroid carcinoma and matched controls
Characteristic Survivors
n = 67
Controls n = 56
P Value
Sex, n (%) 0.878a
Female 58 (86.6) 49 (87.5)
Male 9 (13.4) 7 (12.5)
Age at evaluation, years 34.2 (18.8-61.7) 34.0 (19.4-60.2) 0.431b
Age at diagnosis, years 15.8 (7.9-18.8) - n.a.
Follow-up duration, years 17.8 (5.0-44.7) - n.a.
Nationality, n (%) n.a.
Dutch 67 (100) 56 (100)
Marital status, n (%) 0.129a
Relationship 43 (64.2) 43 (76.8)
No relationship 24 (35.8) 13 (23.2)
Completed education, n (%) 0.128a
Low level 16 (23.9) 6 (10.7)
Medium level 25 (37.3) 21 (37.5)
High level 26 (38.8) 29 (51.8)
Employment, n (%) 0.743a
Employed or full-time student 61 (91.0) 50 (89.3)
Unemployed and no full-time student 6 (9.0) 6 (10.7)
Relationship with survivor, n (%) n.a.
Peer - 48 (85.7)
Sibling - 8 (14.3)
Numbers are shown as median (range). Abbreviation: n.a., not applicable. a Chi-square test. b Mann-Whitney U test.
7
P <0.001). For all other survivor, tumor, treatment, or follow-up characteristics, survivors of the 3 groups did not differ (Supplemental Table 1). Table 1 shows the demographic characteristics of the survivors and their matched controls. Median age of survivors at evaluation was 34.2 years (range 18.8 to 61.7). Most survivors were female (86.6%), were married or in a relationship (64.2%), and had a paid job or were full-time students (91.0%). Tumor, treatment, and follow-up characteristics of survivors are shown in Supplemental Table 1. Survivors had a median follow-up time of 17.8 years (range 5.0 to 44.7) after diagnosis. All survivors underwent a total thyroidectomy and 131I was administered to 97.0%. Of the survivors, 82.1% remained in remission during follow-up. Ten percent had recurrence of the disease and 7% had persistent disease. The 67 survivors gathered 59 controls. Three controls were not included because matching criteria were not met. The survivor and control groups did not differ significantly with respect to sex, age, marital status, employment/education, nationality, or completed education (Table 1). Almost 14% of the controls were siblings of the survivors.
QoL in adult survivors of pediatric DTC (vs. matched controls)
Table 2 summarizes self-reported QoL measurements in survivors and controls. For the subscales of the questionnaires, survivors had significantly lower scores on physical functioning and role limitations due to physical problems as compared with controls (P = 0.031 and P = 0.021, lower scores represent worse HRQoL) [Figure 2(A) and 2(B)].
Mental fatigue scores were significantly higher in survivors (P = 0.012, higher scores represent more fatigue) [Figure 2(C)]. Scores on the remaining SF-36 and MFI-20 subscales and all HADS subscales did not differ significantly between survivors and controls. In 12 out of 15 subscales on generic HRQoL, fatigue, anxiety, and depression, the interquartile range of scores was larger for survivors than for controls; for none
Table 2. Quality of life in adult survivors of pediatric differentiated thyroid carcinoma and matched controls
Physical functioning 95 (85-100)a 100 (95-100)b 0.031c
Role limitations due to physical problems 100 (75-100)a 100 (100-100)b 0.021c
Bodily pain 84 (72-100)a 100 (74-100)d 0.104
Social functioning 88 (63-100) 100 (81-100) 0.065
Mental health 84 (72-92)a 84 (76-92) 0.723
Role limitations due to emotional problems 100 (67-100)a 100 (67-100)b 0.711
General health perceptions 72 (57-87)a 77 (67-89) 0.231
Vitality 65 (50-78)a 70 (60-80) 0.194
Mental component summary scale 54 (41-56)e 53 (47-56)b 0.961
7
Physical component summary scale 53 (47-57)e 57 (53-59)b 0.024c MFI-20
General fatigue 10 (8-15) 9 (5-12) 0.075
Physical fatigue 8 (5-12) 6 (4-10) 0.083
Reduced activity 8 (5-11) 8 (5-11) 0.613
Reduced motivation 6 (4-9) 6 (4-9) 0.879
Mental fatigue 9 (5-15) 7 (4-10) 0.012c
Total 41 (31-57) 36 (27-54) 0.129
HADS
Anxiety 4 (2-9) 3 (2-6) 0.317
Depression 1 (0-4) 1 (0-3) 0.964
Total 6 (2-11) 4 (3-9) 0.392
THYCA-QoL single items, n (%) No A Little Quite a Bit Very Much
Problems with scara 56 (84) 8 (12) 2 (3) 0 (0)
Numbers are shown as median (interquartile range). Abbreviations: SF-36, short form 36; MFI-20, multidimensional fatigue inventory 20; HADS, hospital anxiety and depression scale; and THYCA-QoL; thyroid cancer-specific health-related quality of life. For SF-36, higher scores indicate a better HRQoL; subscales range from 0 to 100. For MFI-20, higher scores indicate more fatigue; subscales range from 4 to 20; the total scale ranges from 20 to 100.
For HADS, higher scores indicate higher levels of anxiety and/or depression; subscales range from 0 to 21; the total scale ranges from 0 to 42. For THYCA-QoL, a higher score indicates more symptoms; subscales range from 0 to 100. Scores on the SF-36, MFI-20, and HADS were compared using the Mann–Whitney U test. a n = 66 because 1 survivor filled in the questionnaire erroneously. b n = 54 because 2 controls filled in the questionnaire erroneously.
c P < 0.05. d n = 53 because 3 controls filled in the questionnaire erroneously. e n = 65 because 2 survivors filled in the questionnaire erroneously.
of the 15 subscales the interquartile range of scores for controls was larger than that for the survivors. For the component scales, scores on the physical component scale were significantly lower in survivors compared to controls (P = 0.024; lower scores
7
7
represent worse HRQoL). Scores of other summary scales (mental component scale, total MFI-20, and total HADS scores) did not differ significantly between survivors and controls. Scores of survivors with a second malignant neoplasm (SMN) are reported in Supplemental Table 2. Three survivors developed a SMN after DTC. Scores of QoL of the 2 women diagnosed with breast cancer were similar compared with the survivors that were not diagnosed with a SMN. For 1 woman diagnosed with a high-grade cervical intraepithelial neoplasia, QoL scores were lower compared with the survivors that were not diagnosed with a SMN after DTC. Siblings of survivors did not differ significantly from nonsiblings in scores on questionnaires (data not shown). For thyroid cancer-specific QoL, scores of survivors are shown in Table 2 and Figure 2(D). Most frequently reported were neuromuscular, throat/mouth, psychological and sensory complaints, feeling chilly, and headache.
Characteristics associated with QoL in adult survivors of pediatric DTC Characteristics that were possibly associated with QoL in adult survivors of pediatric DTC are shown in Supplemental Table 3. Male survivors reported higher levels of reduced motivation and more depression than did females (P = 0.007 and P = 0.009). Unemployment was associated with lower scores on physical functioning and the physical component scale (P = 0.005 and P = 0.003) and higher levels of general fatigue, physical fatigue and total fatigue (P = 0.009, P = 0.009, and, P = 0.004, respectively). Treatment with a higher cumulative dose of 131I was associated with more complaints of headache (P = 0.006). Survivors with recurrent or persistent disease and survivors with a higher level of TSH suppression reported more complaints about the scars in their necks (P <0.001 and P = 0.002). Marital status at follow-up, educational level, age at follow-up, all tumor characteristics (i.e. age at diagnosis, tumor node metastasis stage, and histology), surgical complications (permanent hypoparathyroidism and recurrent laryngeal nerve injury), and follow-up duration showed no significant associations with QoL (HRQoL, fatigue, anxiety, depression, and/or thyroid cancer-specific HRQoL).
Figure 2. Quality of Life (QoL) scores in survivors of pediatric differentiated thyroid carcinoma. An asterisk (*) indicates a statistically significant difference. (A-C) Scores of survivors vs. controls. (A) Physical functioning (Short Form 36, health-related QoL) scores of survivors of vs. matched controls. Scale ranges from 0 to 100. A higher score indicates a better QoL. (B) Limitations in role functioning due to physical health (Short Form 36, health-related QoL) scores of survivors vs. matched controls. Scale ranges from 0 to 100. A higher score indicates a better QoL. (C) Mental fatigue (Multidimensional Fatigue Inventory, fatigue) scores of survivors vs. matched controls. Scale ranges from 4 to 20. A higher score indicates a worse QoL. (D) Scores of survivors only (THYCA-QoL, thyroid cancer-specific health-related QoL).
Scores of survivors on neuromuscular, throat/mouth, psychological, and sensory complaints are shown.
Scales range from 0 to 100. A higher score indicates a worse QoL.
7
DISCUSSION
In the present study, we focus on QoL in long-term adult survivors of pediatric DTC, and our results show overall normal HRQoL, fatigue, anxiety, and depression in these survivors compared to matched controls. This normal QoL is similar to that found in studies performed in survivors of adolescent DTC (8, 9). Although survivors reported more physical problems, role limitations due to physical problems, and mental fatigue, the overall scores of survivors on these domains were still within the normal range.
Thyroid cancer-specific complaints were present in some survivors, but most survivors reported relatively few or no complaints [Table 2 and Figure 2(D)]. Another reflection of the normal QoL of the survivors is their full participation in society, exemplified in most of these survivors by high employment and relationship rates and an educational level comparable to matched controls. Thereby, most (88%) of the survivors reported to have an interest in sex.
Remarkably, the overall QoL in survivors of pediatric DTC seems to be better than the QoL of survivors diagnosed with DTC at older ages. HRQoL, fatigue, anxiety, and depression were reported as worse in both short- and long-term survivor studies in adult DTC compared with controls (10, 11). Additionally, survivors of (young) adult DTC seem to experience more severe thyroid cancer-specific complaints than do survivors of pediatric DTC (5, 13, 22).
Although overall QoL is normal in pediatric DTC survivors compared to controls,
Although overall QoL is normal in pediatric DTC survivors compared to controls,